To use Elasticsearch, customers will need to install a cluster on a physical server that meets certain specifications. We recommend a minimum of 32GB of RAM for optimal performance.
The preferred approach is to install Elasticsearch using Docker containers. The basic architecture consists of one master node with 1GB of RAM, two data nodes with 15GB of RAM each, and one monitoring node with 1GB of RAM.
This setup provides a reliable and scalable solution for managing Elasticsearch data and monitoring its performance.
https://toanmx.notion.site/f5073b6d9382418783d6905e4d3fb3e5?v=def3cce483c2458f8dc9116be66110a3
- Setup Elasticsearch
- Set up 4 nodes: one master-eligible node and two data nodes, one monitoring node
- Specify the node specifications, including CPU, RAM, and storage capacity
- Specify the operating system and Elasticsearch versions
- Specify the backup frequency (e.g. daily) and retention policy
- Consider disaster recovery mechanisms and testing procedures
- Document the setup, configuration, testing, and monitoring procedures
- Elasticsearch Cluster Monitoring
- Setup Kibana and Metricbeat for the monitoring dashboard
- Set up alerts for high CPU usage and other important metrics